Morphological and molecular characterization ofmycorrhizal fungi isolated from neotropicalorchids in Brazil

Olinto Liparini Pereira, Maria Catarina Megumi Kasuya, Arnaldo Chaer Borges,Elza Fernandes de Araújo

Abstract: To initiate a conservation program of the Orchidaceae from the Brazilian Atlantic rain forest with the pur-pose of ex situ conservation or reintroduction in the State of Minas Gerais, seven mycorrhizal Rhizoctonia-like fungalstrains were isolated from roots of seven neotropical orchid species from three different Atlantic rain forest fragments.Taxonomic studies revealed that the isolates belong to the genera Ceratorhiza and Epulorhiza. The Epulorhiza isolateswere identified as Epulorhiza repens (N. Bernard) R.T. Moore and Epulorhiza epiphytica Pereira, Rollemberg etKasuya. RAPD analysis indicated higher polymorphism between Epulorhiza epiphytica and Epulorhiza repens thanfound in the PCR–RFLP analysis. RAPD and morphological analyses indicated a degree of relatedness among theCeratorhiza isolates obtained from the roots of different Oncidium species. A combination of morphological and mo-lecular characterizations permitted integration of fungal strain identification with genetic relatedness among the isolates,thus allowing some inferences to be made on specificity of these endosymbionts under field conditions.

Key words: biodiversity, Ceratorhiza, Epulorhiza, orchid mycorrhiza, Rhizoctonia-like, symbiosis, specificity.

Résumé : Les auteurs travaillent à mettre sur pied un programme de conservation pour les Orchidaceae de la forêt om-brophile atlantique du Brésil, avec l’objectif d’effectuer de la conservation ex situ et de la réintroduction, dans l’état deMinas Gerais. À cette fin, ils ont isolé sept souches de champignons de type Rhizoctonia, à partir de racines de septespèces d’orchidées néotropicales, provenant de trois fragments différents de la forêt ombrophile de l’Atlantique. Lesétudes taxonomiques révèlent que les isolats appartiennent aux genres Ceratorhiza et Epulorhiza. Ils ont identifié lesisolats d’Epulorhiza comme Epulorhiza repens (N. Bernard) R.T. Moore et Epulorhiza epiphytica Pereira, Rollemberget Kasuya. L’analyse RAPD montre un polymorphisme plus grand entre l’Epulorhiza epiphytica et l’Epulorhiza repens,que celui qu’on retrouve avec les analyses PCR–RFLP. Les analyses RAPD et morphologiques indiquent un certain de-gré de relation entre les isolats provenant des Ceratorhiza obtenus des racines de différentes espèces d’Oncidium. Lacombinaison des caractéristiques morphologiques et moléculaires permet d’intégrer l’identification des souches fongi-ques avec la relation génétique entre les isolats, permettant ainsi de déduire des informations sur la spécificité de cesendosymbiontes, sous des conditions de terrain.

Mots clés : biodiversité, Ceratorhiza, Epulorhiza, mycorhizes des orchidées, type Rhizotonia, symbiose, spécificité.

[Traduit par la Rédaction] Pereira et al. 65

Introduction

In nature, orchids utilize mycorrhizal fungi to initiate seedgermination and for seedling establishment, at least in theearly stages of their development (Andersen and Rasmussen1996; Harvais and Hadley 1967; Currah et al. 1997; Petersonet al. 1998; Rasmussen 2002). The fungi provide exogenouscarbon sources, resulting in the orchids dependence on endo-mycorrhizal fungi as an absolute requirement for completingtheir life cycle, in one of the most pervasive mutualistic rela-tionships found in terrestrial ecosystems (Leake 1994; Smith

and Read 1997). The Brazilian Atlantic rain forest is one ofthe most threatened tropical ecosystems in the world (Mori1989). This ecosystem is recognized for its extreme bio-diversity, with a very diverse array of terrestrial and, espe-cially, epiphytic orchids (Miller et al. 1996). However,disturbances caused by land-clearing for agriculture, for-estry, urbanization, pastureland, natural resource and extrac-tion, as well as frequent fires, have resulted in an increasedthreat of extinction for many orchid species in Minas Gerais(Mendonça and Lins 2000) and other Brazilian states(Ruschi 1986; Miller et al. 1996).

Can. J. Bot. 83: 54–65 (2005) doi: 10.1139/B04-151 © 2005 NRC Canada

54

Received 12 February 2004. Published on the NRC Research Press Web site at http://canjbot.nrc.ca on 18 January 2005.

O.L. Pereira,1 M.C.M. Kasuya, and A.C. Borges. Laboratório de Associações Micorrízicas, Departamento de Microbiologia,Instituto de Biotecnologia Aplicada a Agropecuária (BIOAGRO), Universidade Federal de Viçosa, 36571-000 Viçosa/MG, Brazil.E. Fernandes de Araújo. Laboratório de Genética Molecular e de Microrganismos, Departamento de Microbiologia, Instituto deBiotecnologia Aplicada a Agropecuária (BIOAGRO), Universidade Federal de Viçosa, 36571-000 Viçosa/MG, Brazil.

1Corresponding author (e-mail: liparini@bol.com.br).

Mycorrhizal fungi associated with the roots of orchidshave thus been isolated, characterized, and stored as impor-tant resources for the conservation of orchid species (Ardittiet al. 1990; Zettler 1997a, 1997b; Batty et al. 2001a, 2001b).Studies have employed mycorrhizal fungi for symbiotic seedgermination of some endangered orchid species (Zettler andMcInnis 1993; Zettler and Hofer 1998; Zettler et al. 2000;Zettler et al. 2001; Takahashi et al. 2001; Stewart and Zettler2002) with potential to be used to establish orchid popula-tions in restored habitats. However, most work on orchidmycorrhizal fungi has dealt with terrestrial, temperateorchids, and very few reports on associations betweenendomycorrhizal fungi and epiphytic or tropical species existin the literature (Bayman et al. 1997; Richardson and Currah1995; Richardson et al. 1993; Zettler et al. 1998, 1999;Pereira et al. 2003b).

Identifying Rhizoctonia and Rhizoctonia-like endophytesfrom mycorrhizal orchid roots is very problematic becauseof a paucity of sufficiently distinctive and stable morpholog-ical features in culture (Andersen 1990). However, a varietyof approaches have been developed to identify them: mor-phological characters (Currah and Zelmer 1992; Currah andSherburne 1992; Andersen 1996; Currah et al. 1997), anasto-mosis groups (Carling et al. 1999; Sen et al. 1999), and theuse of molecular techniques (Sen et al. 1999; Kristiansen etal. 2001; Shan et al. 2002; Otero et al. 2002; Ma et al.2003).

The objective of our study was to characterizeRhizoctonia-like mycorrhizal fungi associated with sevenneotropical orchid species native to Minas Gerais State,Brazil, by employing morphological and molecular tech-niques.

Materials and methods

Collection of orchid root samplesIntact young and healthy roots of seven neotropical orchid

species from Atlantic rain forest fragments in Minas GeraisState, Brazil, were selected for isolation of orchidaceousendosymbiotic fungi (Table 1). Root fragment samples fromabout five plants of each species were collected in their na-tive habitat between 2000 and 2001, transported within 1 hto the laboratory, and washed under running tap water. Sam-ples were collected only when the plants were flowering tofacilitate identification of the orchid species. Each sampleconsisted of about 15–20 roots fragments of 5–10 cm each.Part of the root sample was selected for anatomical studiesand the other part for endosymbiont isolation.

Isolation and morphological characterizationSamples of healthy root fragments from epiphytic orchid

species were surface-sterilized by immersion for 1 min in a70% ethanol solution, followed by immersion for 5 min in2% sodium hypochloride solution and finally by rinsing fivetimes with sterilized distilled water. The roots were thendecorticated with a sterile scalpel, macerated with a pestle ina porcelain mortar, spread on Petri dishes containing 25 mLof modified Melin-Norkrans medium (Marx 1969), and in-cubated at 26 °C in the dark. Pelotons that were visible onthe isolation medium surface under the light microscopewere monitored for the presence of active hyphal growth.

Hyphal tips were selected, removed aseptically with ascalpel, and transferred into Petri dishes containing potatodextrose agar (PDA; Difco Laboratories, Detroit, Michigan),and incubated for 7 d at 26 °C for the isolation of onlypeloton-forming fungi (Pereira et al. 2003b). Agar plugs(9 mm in diameter) from the border of resulting colonieswere transferred into Petri dishes containing either PDA,corn meal agar (CMA, Difco), coconut milk agar, malt ex-tract agar, or oat meal agar (Zelmer and Currah 1995) formorphological characterization. Colony diameter, colour,border, and aerial mycelial aspect were recorded for fivereplicate plates after 72 h of incubation in the dark at 26 °C.Monilioid cells were examined in colonies grown on CMA,and runner hyphae were measured in colonies grown onPDA. Nuclei numbers were counted in young hyphal cellsafter staining with HCl-Giemsa (Saksena 1961). Cultureswere kept on PDA or in sterilized distilled water at 4 °C forlong-term storage (Sneh and Adams 1996).

Enzymatic assaysIsolates were grown on minimal medium with carboxy-

methyl cellulose as the sole carbon source for cellulase(Teather and Wood 1982). After 7–10 d of incubation at26 °C, the agar medium was flooded with an aqueous solu-tion of Congo red (1 mg·mL–1) for 15 min. The Congo redsolution was then poured off, and the medium was treatedfurther by flooding with 1 mol·L–1 NaCl for 15 min. Thecellulose hydrolysis zone was readily observable on the me-dium surface. Polyphenol oxidase activity was detected us-ing the method of Davidson et al. (1938).

Identification of fungal speciesThe isolated fungi were identified using a dichotomous

key for genera of mycorrhizal fungi associated with orchids(Currah and Zelmer 1992), and a key to Epulorhiza species(Currah et al. 1997) that includes Epulorhiza repens (N. Ber-nard) R.T. Moore (Moore 1987, 1996), Epulorhiza anaticula(Currah) Currah (Currah et al. 1990), Epulorhiza alber-taensis Currah and Zelmer (Currah and Zelmer 1992),Epulorhiza calendulina Zelmer and Currah (Zelmer andCurrah 1995), and Epulorhiza inquilina Currah, Zettler, andMcInnis (Currah et al. 1997). Specimen subcultures arebeing maintained in the Departamento de Microbiologia /BIOAGRO (see Table 2 for isolate codes) and voucher speci-mens were deposited at VIC Herbarium. Random amplifiedpolymorphic DNA (RAPD) and polymerase chain reaction –restriction fragment length polymorphism (PCR–RFLP)analyses of the rDNA internal transcribed spacer (ITS)region were performed to supplement the limited morpho-logical characteristics available for Rhizoctonia-like fungiidentification.

Teleomorph induction experimentsInduction of basidiome production in the seven isolates

was attempted by the soil-over-culture (Tu et al. 1969) andplant infection (Flentje 1956) techniques. Eucalyptusgrandis W. Hill ex Maiden and Lycopersicum esculentumMill. seedlings were used for infection. The induction wasalso carried out in Petri dishes containing 25 mL of eachmedia: PDA, CMA, coconut milk agar, malt extract agar, oatmeal agar (Zelmer and Currah 1995), V-8 agar juice

© 2005 NRC Canada

Pereira et al. 55

(Dhingra and Sinclair 1995), and vegetable–broth–agar me-dium (Pereira et al. 2003a). Cultures were incubated in thedark for 48 h and later submitted to photoperiods of 12 hnear-ultraviolet irradiation : 12 h darkness (Leach 1962) for25 d at 25 °C. The experiments were routinely observed forfruit body formation. Each treatment was replicated threetimes and the experiment was carried out twice.

Microscopic observationsRoots collected from wild orchids were transversally cut

using a freezing-stage microtome in 30-µm-thick slices thatwere mounted on glass slides and examined using an lightmicroscope for peloton presence. Pelotons were also visual-ized on the Melin-Norkrans agar surface using an invertedmicroscope during isolation.

DNA extractionPrior to DNA extraction, the isolates were grown on cello-

phane membranes overlaid on PDA for 2 weeks in the darkat 26 °C. The outer borders of young (7 d old) colonies wereexcised and the underlying cellophane was removed. TotalDNA was extracted from 0.8 to 1.0 g of the fresh mycelium,according to the procedures of Speacht et al. (1982). TheDNA pellet was dissolved in 30 µL of double-distilled waterand the DNA concentration was estimated by comparisonwith known standards in 0.8% agarose gels stained withethidium bromide. DNA samples were stored in Eppendorftubes at –20 °C.

RAPD analysisPCR amplification of DNA sequences was performed with

each of the arbitrarily selected primers: OPH 06, OPD 04,OPG 11, OPF 14, OPH 03, OPP 01, OPG 05, OPE 02, OPC02, OPC 05, OPC 08, OPC 13, OPJ 02, OPW 07, OPH 14,OPW 02, and OPE 19 (Operon Technologies Inc., Alameda,California). Each 20-µL PCR reaction contained 10 mmol·L–1

Tris-HCl (pH 8.3), 50 mmol·L–1 KCl, 2 mmol·L–1 MgCl2,100 µmol·L–1 each of dATP, dCTP, dGTP, and dTTP,0.4 µmol·L–1 of one random primer, 10 ng of genomic DNA,and 1 U of Taq DNA polymerase. Negative controls withoutDNA were run in all experiments. Amplification was per-formed in a PTC-100 Programmable Thermal Controller(MJ Research, Inc., Watertown, Massachusetts), pro-grammed for 40 cycles, with each cycle consisting of onedenaturating step at 94 °C for 60 s, one annealing step at50 °C for 60 s, and one extension step at 72 °C for 90 s. Af-ter the 40th cycle, a final extension step was performedat 72 °C for 7 min. Amplification products were electro-

phoresed in 1.5% agarose gels immersed in TEB runningbuffer (90 mM Tris-borate, 1 mmol·L–1 EDTA, pH 8.0) andrun for 4 h at 100 V. The gels were stained with ethidiumbromide (0.5 µg mL–1) and photographed under UV lightwith a Polaroid camera. The images were also captured andstored using a photodocumentation system (Eagle Eye II,Stratagene, La Jolla, California).

PCR amplification and RFLP analysisPrimers ITS1 (5′-TCCGTAGGTGAACCTGCGG-3′) and

ITS4 (5′-TCCTCCGCTTATTGATATGC-3′), constructed formolecular phylogenetic studies (White et al. 1990), wereused for amplification of the fungal rDNA ITS region of allisolates. The components for 25-µL PCR reactions were10 ng of total DNA, 40 pmol of each primer, 10 mmol·L–1

Tris-HCl (pH 8.3), 50 mmol·L–1 KCl, 2 mmol·L–1 MgCl2,0.1 mmol·L–1 of each dNTP (dATP, dCTP, dGTP, anddTTP), and 1 U of Taq DNA polymerase. The cycling pa-rameters were 40 cycles, each cycle consisting of a denatur-ation step at 94 °C for 60 s, an annealing step at 52 °C for60 s, and an extension step at 72 °C for 90 s. After the 40thcycle, a final extension step at 72 °C for 7 min was per-formed. The amplification reaction was performed in a PTC-100 thermocycler (MJ Research, Inc.). After amplification,the DNA products were analysed by eletrophoresis in a 1.5%(w/v) agarose gel immersed in TBE buffer (90 mmol·L–1

Tris-borate, 2 mmol·L–1 EDTA, pH 8.0) or precipitated forRFLP analysis. DNA was precipitated by adding ammoniumacetate to a final concentration of 2.5 mol·L–1 followed by2.5 volumes of absolute ethanol. The samples were kept at –20 °C for 5 h and centrifuged at 14 000g for 30 min. Pelletswere washed with 70% (v/v) ethanol, resuspended in 10 µLH2O, and digested with five arbitrarily selected endo-nucleases (AluI, HaeIII, HinfI, EcoRI, and ClaI) in a 20-µLfinal volume. DNA digestion was performed following themanufacturer’s instructions. DNA fragments were size-fractionated in 2% (w/v) agarose gel. The gels were stainedwith ethidium bromide (0.5 µg·mL–1), and photographed un-der UV light with a Polaroid camera. The images were alsocaptured and stored using the photodocumentation systemEagle Eye II.

Numerical analysisQualitative morphological features were used to define

morphological groups. The selected morphological charac-teristics were based on cultural features known to be of im-portant taxonomical value for mycorrhizal Rhizoctonia-likefungi (Currah and Zelmer 1992). The scoring was done con-

© 2005 NRC Canada

56 Can. J. Bot. Vol. 83, 2005

Orchid species Habitat Location* Isolate code

Epidendrum rigidum Jacq. Epiphytical Pedra do Anta M1Oncidium flexuosum (Kunth) Lindl. Epiphytical Viçosa M2Isochilus lineares (Jacq.) R. Br. Epiphytical Carangola M3Maxillaria marginata Fenzl Epiphytical Carangola M4Oeceoclades maculata (Lindl.) Lindl. Terrestrial Viçosa M5Polystachya concreta (Jacq.) Garay and Sweet Epiphytical Pedra do Anta M6Oncidium varicosum Lindl. and Paxton Epiphytical Viçosa M7

*Cities from Minas Gerais State, Brazil.

Table 1. Orchid species, habitat, and location from which Rhizoctonia-like endophyteswere isolated and used in our study.

sidering the presence (1) or absence (0) of such qualitativemorphological features.

Scoring of PCR–RFLP and RAPD analyses was doneconsidering presence (1) or absence (0) of determined DNAfragments for different samples. The presence of a deter-mined band (similar size) in all genotypes compared indi-cated similarity, while presence in one and absence in theother indicated dissimilarity. The data were analyzed by thestatistical program GENES (Cruz 1997). The genetic dis-tances among the isolates were calculated using Nei–Li’scoefficient (Nei and Li 1979), according to the formula:

Sii′ = 2a/(a + b + c)

where Sii′ is the Nei–Li’s coefficient for genotypes i and ′i , ais the band or the same allelic form present in genotypes iand ′i , b is the band or the same allelic form present in geno-type i, and c is the band or the same allelic form present ingenotype ′i .

Cluster analyses of the morphological and molecular datawere performed by the unweighted pair-groups method algo-rithm (UPGMA), using the STATISTICA program, version4.2 (Statsoft 1995). To aid the characterization of the groups,the Tocher cluster method and rPearson method, as de-scribed by Cruz and Regazzi (1994) were also performed.Agreement between values in the three matrices was esti-mated from the coincidence coefficient, which represents thecorrelation between the highest and lowest genetic distancevalues, starting from the average value of each matrix (Cruzand Regazzi 1994).

Results

Peloton observationThe majority of root sections showed intense colonization.

Intact (Figs 1, 2) or degraded (Fig. 3) pelotons were ob-served in about 75%–80% of the root cortical cells ofIsochilus lineares, Maxillaria marginata, Oncidiumflexuosum, Oncidium varicosum, and Oeceoclades maculata,whereas Epidendrum rigidum and Polystachya concretashowed very low colonization frequency, with only about5%–10% of the root cortical cells colonized. In intenselycolonized species, the large cortical cells in the middle re-gions of the root sections contained clumps of degeneratinghyphae, whereas cells at the edge contained intact hyphae.Only one distinct peloton-forming fungus was found in allroot segment samples of a given species. Intact pelotonsof Epidendrum rigidum, P. concreta, and Oeceocladesmaculata were formed by narrower hyphae, whereas intactpelotons of the remaining species were formed by broaderhyphae. Fungal hyphae were observed on the velamen sur-face of root fragments of all epiphytic species examined inour experiment.

IsolationIt was much more difficult to isolate endosymbiotic fungi

from epiphytic orchid species than from the terrestrial or-chid, Oeceoclades maculata. Dark septate endophytes(Jumpponen and Trappe 1998), were consistently isolatedfrom the culture media of both types of orchid species. Allisolates from a given orchid species appeared to be the samefungal species, based on morphological and morphometric

data (not shown). Therefore, one isolate was selected foreach plant species and codified, in order of isolation, as M1to M7, corresponding to the host species Epidendrumrigidum, Oncidium flexuosum, Isochilus lineares,M. marginata, Oeceoclades maculata, P. concreta, andOncidium varicosum, respectively (Table 1).

Cultural characteristics, nuclear condition, enzymaticassays, and isolate identification

The morphological characteristics analyzed and the resultsof the enzymatic tests are listed in Table 2. All isolates werebinucleate (Fig. 4) and formed monilioid cells in culture(Figs. 5–8). Although all isolates were binucleate, two dis-tinct groups could be recognized, based on morphology,growth on culture media, and enzymatic tests (Fig. 9). Thefirst group, including isolates M1, M5, and M6, consists offungi with scant and submerged mycelia, with very slowgrowth in culture, thin vegetative hyphae, mostly sphericalmonilioid cells, and absence of polyphenol oxidase activiy.The second group, containing isolates M2, M3, M4, and M7,consists of fungi with abundant, cottony aerial mycelia, highgrowth rates in culture, broad vegetative hyphae, mostly cy-lindrical to barrel-shaped monilioid cells, and presence ofpolyphenol oxidase activity. No clamped hyphae were ob-served and pelotons formed by hyphae with clamps were notnoted.

All peloton-forming isolates were Rhizoctonia-like fungibased on morphological criteria. Isolates M1, M5, and M6belong to the genus Epulorhiza and isolates M2, M3, M4,and M7 belong to Ceratorhiza. Epulorhiza isolates wereidentified at the species level, Epulorhiza repens (N. Ber-nard) R.T. Moore and Epulorhiza epiphytica Pereira, Rol-lemberg et Kasuya (Pereira et al. 2003b) (Table 2), whilespecies identification was not possible for the Ceratorhizaisolates because the morphological features of these organ-isms were not sufficiently distinctive and stable in culture.No Ceratorhiza isolate was assignable to C. pernacatenaZelmer & Currah (Zelmer and Currah 1995).

Teleomorph inductionInduction of basidiome production was unsuccessful in all

seven isolates.

PCR–RFLP analysisThe ITS region, targeted by the universal primers ITS1

and ITS4 was amplified in all seven isolates (Fig. 10). PCRproducts varied in size from 640 to 730 bp. Amplification ofITS sequences from Epulorhiza isolates M1, M5, and M6generated a common fragment of approximately 640 bp.Fragments from Ceratorhiza isolates M2, M3, M4, and M7ranged from 660 to 730 bp (Table 3). The longest fragment(730 bp) was from the sole sclerotia-forming Ceratorhizaisolate (M7). The smaller polymorphic fragment products,produced by digestion of the ITS regions with the fiveendonucleases, AluI, HaeIII, EcoRI, HinfI, and ClaI, areshown in Table 3. Endonucleases AluI, EcoRI, HinfI, andClaI generated the same restriction pattern for all Epulorhizaisolates while HaeIII generated different restriction patterns,separating Epulorhiza repens (M5) from both Epulorhizaepiphytica isolates (M1 and M6). The pair of Epulorhizaepiphytica isolates M1 and M6 produced the same restric-

© 2005 NRC Canada

Pereira et al. 57

tion patterns for all five endonucleases utilized in the PCR–RFLP analysis, as did the pair of Ceratorhiza isolates M3and M4. All endonucleases generated different restriction

patterns for the M2 and M7 and the M3 and M4 pairs ofCeratorhiza isolates (Table 3). A genetic distance matrixwas constructed (not shown) based on restriction products

© 2005 NRC Canada

58 Can. J. Bot. Vol. 83, 2005

Figs. 1 and 2. Intact pelotons observed in root cortical cells of two neotropical orchid species. Fig. 1. Isochillus lineares. Scale bar =10 µm. Fig. 2. Oncidium varicosum. Scale bar = 10 µm. Fig. 3. Digested peloton observed in the root cortical cell of Polystachiaconcreta. Scale bar = 10 µm. Fig. 4. Binucleate nuclear condition of young hyphae cell of Epulorhiza repens, stained with HCl-Giemsa. Scale bar = 10 µm. Figs. 5–8. Monilioid cells formed in corn meal agar by four Rhizoctonia-like endophytes. Fig. 5.Epulorhiza repens isolate M5. Scale bar = 20 µm. Fig. 6. Epulorhiza epiphytica isolate M6. Scale bar = 20 µm. Fig. 7. Ceratorhiza sp.isolate M7. Scale bar = 30 µm. Fig. 8. Ceratorhiza sp. isolate M3. Scale bar = 30 µm.

© 2005 NRC Canada

Pereira et al. 59

Isol

ate

code

Cha

ract

eris

tics

M1

M2

M3

M4

M5

M6

M7

Col

ony

onP

DA

Col

orC

ream

Pale

brow

nW

hite

Whi

teC

ream

Cre

amPa

lebr

own

App

eara

nce

Seb

aceo

usC

otto

nyC

otto

nyC

otto

nyS

ebac

eous

Seb

aceo

usC

otto

nyM

argi

nS

ubm

erge

dA

eria

lA

eria

lA

eria

lS

ubm

erge

dS

ubm

erge

dA

eria

l

Aer

ial

myc

elia

onP

DA

App

eara

nce

Sca

ntA

bund

ant

Abu

ndan

tA

bund

ant

Sca

ntS

cant

Abu

ndan

tD

iam

eter

ofhy

phae

( µm

)2

5.0–

6.0

4.0–

5.0

4.0–

5.0

2.0–

3.0

25.

5–6.

5N

ucle

arco

ndit

ion

Bin

ucle

ate

Bin

ucle

ate

Bin

ucle

ate

Bin

ucle

ate

Bin

ucle

ate

Bin

ucle

ate

Bin

ucle

ate

Mon

ilio

idce

lls

onC

MA

Dim

ensi

ons

( µm

)(7

.0–1

0)×

(8.0

–9.5

)(2

0–26

)×

(8.0

–10)

(17–

25)

×(9

.5–1

4)(1

7–27

)×

(9.5

–14)

(14–

16)

×(1

0–15

)(7

.0–1

0)×

(8.0

–9.5

)(1

8–24

)×

(7.5

–10)

Sha

peG

lobo

seB

arre

lB

arre

lB

arre

lG

lobo

seto

elli

psoi

dal

Glo

bose

Bar

rel

Sur

face

Fove

ate

Sm

ooth

Sm

ooth

Sm

ooth

Sm

ooth

Fove

ate

Sm

ooth

Scl

erot

ia**

Abs

ent

Abs

ent

Abs

ent

Abs

ent

Abs

ent

Abs

ent

Pre

sent

Tele

omor

phA

bsen

tA

bsen

tA

bsen

tA

bsen

tA

bsen

tA

bsen

tA

bsen

t

Enz

imat

icas

say

For

cell

ulas

eP

rese

ntP

rese

ntP

rese

ntP

rese

ntP

rese

ntP

rese

ntP

rese

ntFo

rpo

life

nol-

oxid

ase

Not

dete

ctab

leP

rese

ntP

rese

ntP

rese

ntN

otde

tect

able

Not

dete

ctab

leP

rese

nt

Iden

tifi

cati

onE

pulo

rhiz

aep

iphy

tica

Cer

ator

hiza

sp.

Cer

ator

hiza

sp.

Cer

ator

hiza

sp.

Epu

lorh

iza

repe

nsE

pulo

rhiz

aep

iphy

tica

Cer

ator

hiza

sp.

Vou

cher

VIC

2780

2V

IC27

803

VIC

2780

4V

IC27

805

VIC

2780

6V

IC27

807

VIC

2780

8

Not

e:PD

A,

pota

tode

xtro

seag

ar;

CM

A,

corn

mea

lag

ar.

Tab

le2.

Sum

mar

yof

cult

ure

and

mor

phom

etri

cch

arac

teri

stic

s,an

den

zim

atic

assa

yre

sult

sfo

rth

ese

ven

Rhi

zoct

onia

-lik

eis

olat

es.

from the seven rDNA regions analysed. Cluster analysisbased on these genetic distances separated the isolates intothree groups: group I, containing Epulorhiza isolates; groupII, containing the Ceratorhiza isolate M2; and group III,containing Ceratorhiza isolates M3, M4, and M7 (Fig. 11).

RAPD analysisPCR reactions with 17 primers produced a total of 432

reproducible bands. No bands were present in negative con-trols, and most fragments considered for analysis were poly-morphic. Figure 12 shows the amplification pattern obtainedwith primers OPH 03, OPH 06, and OPP 01. The number offragments per primer varied between 18 and 35, with an av-erage of 25. As in the PCR–RFLP analysis, RAPD analysisresulted in the same patterns for Epulorhiza isolates M1 andM6 and Ceratorhiza isolates M3 and M4. Cluster analysisbased on pairwise genetic distances among isolates groupedthem into three groups: group I, containing the Epulorhizaisolates; group II, containing Ceratorhiza isolates M3 andM4; and group III, containing Ceratorhiza isolates M2 andM7 (Fig. 13).

Correlation and coincidence analyses betweenmorphological and molecular characterizations

A simple coincidence analysis was performed using the

genetic distance values obtained between isolate pairs thatrevealed high superior and inferior coincidence percentages(Table 4). Thus, a high degree of similarity is observedamong the groups formed by the different analyses (morpho-logical, PCR–RFLP of the rDNA ITS region, and RAPD)(Table 5).

Discussion

Root colonizationAll mycorrhizas of epiphytic orchids were very similar,

possessing intact pelotons in the outermost layer of corticalcells, as described by Zelmer et al. (1996) for Corallorhizaspecies. These pelotons probably serve as a source of ino-culum for the recolonization of cells deeper in the root cor-tex (Zelmer et al. 1996), and they may also serve as animportant inoculum source for colonization of seeds on ornear the roots of the mother plant. We have observed thatprotocorms, as well as seedlings, are commonly present onthe root surface of mature epiphytic orchid species in thefield, as reported by Miller and Warren (1995) and Milleret al. (1996). Thus, the presence of young pelotons in theborder may be an important reproductive strategy used byepiphytic orchids in their natural environments. Patterns con-sisting of synchronously formed and degraded patches ofpelotons (Zelmer et al. 1996) were not observed in our ex-periment, probably because the root fragments of each plantspecies used in our study were colonized by only one fungalspecies. The isolation of morphologically identical fungifrom each orchid species also supports this hypothesis.

Isolation, identification, and morphologicalcharacterization

Epulorhiza epiphytica from the roots of Epidendrumrigidum and P. concreta was the first mycorrhizal fungi iso-lated from a Brazilian neotropical species (Pereira et al.2003b). The same plant population, from the same geo-graphical region was selected for the present study. A sea-sonal search for a Tulasnella species was conducted at thissite; however, no hymenia were observed. Failure to inducebasidiome production in Epulorhiza epiphytica and the otherisolates was expected. Hymenia of teleomorphs connected toRhizoctonia-like fungi are very rarely observed in culturemedia, and basidiome induction in Ceratorhiza andEpulorhiza is often unsuccessful (Zelmer et al. 1996; Shan

© 2005 NRC Canada

60 Can. J. Bot. Vol. 83, 2005

Fig 9. Cluster analysis of the seven Rhizoctonia-like isolatesbased on qualitative morphological data. The dendrogram wasgenerated from phenotypic similarity coefficients based on theunweighted pair-group method using the arithmetic average.Identification of the isolates was according to Table 2.

Length of digestion fragments (bp) generated by restriction enzymes

IsolateTotal length ofPCR products (bp) AluI HaeIII EcoRI HinfI ClaI

M1 640 600 600 640 180, 150, 130, 90* 610M2 660 420, 110, 80* 370, 110*, 80* 360, 300 350, 300 380, 240M3 660 450, 190 350, 110,* 100* 370, 290 350, 300 410, 230M4 660 450, 190 350 110,* 100* 370, 290 350, 300 410, 230M5 640 600 490, 100* 640 180, 150, 130, 90* 610M6 640 600 600 640 180, 150, 130, 90* 610M7 730 280, 190, 120 540, 120, 110 380, 330 360, 350 400, 280

Note: Identification of the isolates is according to Table 2.*Possible doublet.

Table 3. Fragment lengths produced by each isolate after PCR–RFLP of the rDNA ITS region, followed bydigestion with the restriction enzymes AluI, HaeIII, EcoRI, HinfI, and ClaI.

et al. 2002; Ma et al. 2003). It appears to be easier to inducethe sexual stage in mycorrhizal multinucleate Rhizoctoniain culture (Currah 1986; Zelmer et al. 1996). Epulorhizaepiphytica and Epulorhiza repens were identified in ourstudy since they possess characteristic monilioid cells. How-ever, the Ceratorhiza isolates possess extremely variable,growth-medium-dependent morphological characters, whichhave little taxonomical value for species identification(Andersen 1990).

Epulorhiza repens, anamorph of Tulasnella deliquescens(Juel) Juel, isolated from the mycorrhizas of Oeceocladesmaculata, is reported for the first time in Brazil. Otero etal. (2002) also tried to isolate the mycorrhizal partner ofOeceoclades maculata, but the isolates obtained by themwere too divergent from the other Rhizoctonia-like speciesto be included in their phylogenetic analysis. Epulorhizarepens is a widespread fungal species (Roberts 1999), andthe ability of Oeceoclades maculata to associate with thisfungus may be contributing to the dispersal success of thisorchid species, regarded as a invasive weed in some situa-tions.

Epulorhiza epiphytica was consistently isolated fromroots of Epidendrum rigidum. Epulorhiza has rarely been

documented in epiphytic orchids, although an Epulorhiza sp.isolate was found associated with roots of the epiphytic or-chid Epidendrum conopseum R. Brown (Zettler et al. 1998).Additionally Epulorhiza have been isolated from otherEpidendrum spp. in Brazil.

Based on the cultural characteristics of Epulorhizaepiphytica, this species could be considered congeneric withEpulorhiza repens, Epulorhiza anaticula, and Epulorhizaalbertensis, belonging to the group 1 proposed by Ma et al.(2003) in the molecular phylogeny of Epulorhiza isolatesfrom tropical orchids in Singapore.

Isolating Rhizoctonia-like fungi from the mycorrhizas ofthe six neotropical epiphytic orchids studied is probably dif-ficult because epiphytic orchid species do not have the mas-sive mycorrhizal colonization observed in terrestrial orchids.Little mycorrhizal infection has also been observed for othertropical epiphytic orchid species (Otero et al. 2002).

Only binucleate Rhizoctonia-like fungi were isolated fromthe six epiphytic orchid species in this study. In contrast,in a study by Bayman et al. (1997) only multinucleateRhizoctonia fungi were isolated from roots of epiphytic or-chids, perhaps because of host-specificity, since that studyinvolved only orchids of the genus Lepanthes Swartz. Anti-biotics were also used in the isolation medium by Bayman etal. (1997), so the mycorrhizal fungal diversity associated toLepanthes spp. may have been underestimated, since somemycorrhizal Rhizoctonia-like fungi are sensitive to antibiot-ics in the isolation medium (Zelmer et al. 1996).

The combination of nuclear condition, mycelial morpho-logical characteristics in culture media, and enzyme activi-ties proved useful for identifying orchid mycorrhizalendosymbionts isolated in our study. Cluster analysis ofthese combined characteristics identified three groups(Fig. 9). The first group contained the Epulorhiza isolates,indicating that the analysis clearly separated the fungi at thegenus level. However, only morphometric data and someunique morphological characters, which could not be uti-lized in the cluster analysis, could separate the species with-in Epulorhiza. The second group contained two Ceratorhizaisolates, obtained from different orchid species of the samegenus, sampled in the same geographical location, indicatingthe same or possibly a closely related Ceratorhiza species.

© 2005 NRC Canada

Pereira et al. 61

Fig 10. Fragments produced by PCR–RFLP of the rDNA ITS region (A) and the fragments produced after restriction with the enzymeEcoRI (B). Lane Y, molecular fragment size marker ØX 174 digested with HaeIII. Identification of the isolates was according to Ta-ble 2.

Fig 11. Cluster analysis of the seven Rhizoctonia-like isolatesbased on results of PCR–RFLP of the rDNA ITS region. Thedendrogram was generated from genetic similarity coefficientsbased on the unweighted pair-group method using the arithmeticaverage. Identification of the isolates was according to Table 2.

The third group contained two other Ceratorhiza isolatesobtained from different orchid genera, but from the samegeographical region, suggesting the same nonspecificCeratorhiza species.

Molecular characterization by PCR–RFLP analysisMolecular characterization of multinucleate Rhizoctonia

spp. and binucleate Rhizoctonia-like fungi is commonly usedto infer taxonomic relationships and patterns of genetic di-versity among species (Cubeta et al. 1996; Mordue et al.1996). In our study, cluster analysis based on the resultsof PCR–RFLP separated the isolates into three groups(Fig. 11). The first group was the same as that found by themorphological analysis containing the Epulorhiza isolates.However, this analysis distinguished Epulorhiza repens fromboth Epulorhiza epiphytica isolates, with a genetic distance

of 18% between the two species. Only the restriction patterngenerated by HaeIII revealed polymorphism among bothEpulorhiza species.

PCR–RFLP has been used to clarify the structure ofmultinucleate Rhizoctonia populations by detecting geneticvariations among species (Cubeta et al. 1996; Cubeta andVilgalys 1997), although in some cases, the use of few re-striction enzymes does not allow detection of polymorphism(Pascual et al. 2000). Binucleate Rhizoctonia-like fungi areknown to be a diverse group of organisms, composed ofmany species (Cubeta et al. 1991), and our results emphasizethe need to use the largest possible number of endonucleasesto generate greater polymorphism in this group of organ-isms.

Genetic distance between Epulorhiza epiphytica isolatesM1 and M6, determined by rDNA PCR–RFLP analysis, wasnull, which is in agreement with the phenotypic distancebased on morphological characterization.

At a genetic distance of 90%, all four Ceratorhiza isolateswere placed in one group. However, at a distance of 88%–89%, isolate M2 was separated from the others. Ceratorhizaisolates M2 and M7, closely related by morphological analy-sis, were distinguished by PCR–RFLP analysis. Basedon these results, these isolates appear to be differentCeratorhiza species, as suggested by the different sizes oftheir amplified ITS regions (Table 3). The pair ofCeratorhiza isolates M3 and M4, closely related by morpho-logical analysis, showed the same restriction pattern with allfive endonucleases used, as well as the same amplified ITSregion sizes (Table 3), indicating that both isolates belong tothe same Ceratorhiza species.

Molecular characterization by RAPD analysisRAPD-PCR analysis has already been successfully em-

ployed in the analysis of genetic variation of Rhizoctoniasolani (Duncan et al. 1993; Yang et al. 1996) as well as

© 2005 NRC Canada

62 Can. J. Bot. Vol. 83, 2005

Fig 12. Gel electrophoresis of random amplified polymorphic DNA (RAPD) fragments obtained with the primers OPH03 (A), OPH06(B), and OPP01 (C). Lane Y, molecular fragment size marker λ digested with EcoRI, BamHI, and HindIII. Identification of the isolateswas according to Table 2.

Fig 13. Cluster analysis of the seven Rhizoctonia-like isolatesbased on RAPD–PCR data. The dendrogram was generated fromgenetic similarity coefficients based on the unweighted pair-group method using the arithmetic average. Identification of theisolates was according to Table 2.

mycorrhizal binucleate Rhizoctonia-like fungi (Shan et al.2002).

As in the PCR–RFLP analysis, isolates M1 and M6 ofEpulorhiza and M3 and M4 (Fig. 11) of Ceratorhiza showedthe same patterns in the RAPD analysis (Fig. 12). For bothgenera, these results suggest that each pair contains a uniquenonspecific isolate, since they were obtained from pairs oforchids belonging to different subtribes. Shan et al. (2002)isolated and characterized 21 Rhizoctonia-like fungal strainsfrom four terrestrial orchid species from various Hong Konglocations utilizing morphological and molecular tools andobserved that highly similar isolates were isolated from thesame host species and locations. In our study, the Epulorhizapair M1 and M6 isolated from Epidendrum rigidum andP. concreta, respectively, was collected in Pedra do Anta,and the Ceratorhiza pair M3 and M4 isolated from Isochiluslineares and M. marginata, respectively, was collected inCarangola. These results suggest a great geographic influ-ence similar to that observed by Otero et al. (2002) for someRhizoctonia-like fungi associated with tropical epiphytic or-chids.

Polymorphism among Epulorhiza spp. belonging to groupI was more evident in the RAPD analysis than in the PCR–RFLP analysis. RAPD analyzes the entire fungal genome,whereas PCR–RFLP analyzes a specific region of the ge-nome. Since RAPD analyzes multiple loci per each primer,genetic variations were more readily detected.

RAPD analysis also detected broad genetic variationamong Ceratorhiza isolates M2 and M7; however, at a ge-netic distance of 75% both isolates were grouped together

(Fig. 13). Both isolates seem to have some degree ofrelatedness that was also detected by morphological analy-sis, perhaps because they were isolated from the same geo-graphic region and same host genus. In contrast, they appearto be different species and a possible variation in specificitywithin the single genus Oncidium can exist, similar to thatobserved for the tropical epiphytic orchids Ionopsisutricularioides (Sw.) Lindl. and Ionopsis satyrioides (Sw.)Rchb. f. (Otero et al. 2002).

Conclusion

The most common fungi associated with Brazilian neo-tropical orchids belong to two genera: Epulorhiza andCeratorhiza. Both morphological and molecular approacheswere useful and consistent for grouping the isolates. Con-sidering the low number of plant species and the restrictedgeographical area investigated, more studies are needed toevaluate the biodiversity and specificity of mycorrhizal fungiassociated with Brazilian orchids and to compare their bio-diversity and specificity to that of isolates from other coun-tries.

The Rhizoctonia-like fungi are taxonomically problematicbecause they consist of a complex of anamorphic basidio-mycetes belonging to different genera and they lackconidiogenesis in their asexual stage (Mycelia Sterilia).These facts complicate mycologists’ understanding of theirecology and global diversity. The integration of differentmethods and techniques should lead to the identification of

© 2005 NRC Canada

Pereira et al. 63

Variables Data No.

Superiorcoincidence(%)

Inferiorcoincidence(%) rPearson

Morphology × PCR–RFLP 19 80 80 0.7700**Morphology × RAPD 19 90 90 0.6693**PCR–RFLP × RAPD 19 90 70 0.8303**

**Statistical significance at p < 0.01 according to a t test.

Table 4. Coincidence analysis based on qualitative and quantitative data from morphologi-cal characterization, PCR–RFLP of the rDNA ITS region, and RAPD analyses for theseven orchid mycorrhizal fungal isolates.

Molecularcharacterization*

IsolateMorphologicalcharacterization* PCR–RFLP RAPD Identification

M1 a a a Epulorhiza epiphyticaM2 b b b Ceratorhiza sp.M3 c c c Ceratorhiza sp.M4 c c c Ceratorhiza sp.M5 a a a Epulorhiza repensM6 a a a Epulorhiza epiphyticaM7 b c b Ceratorhiza sp.

*The same letter within a column indicates that the means are not significantly different according tothe Tocher group analysis.

Table 5. Comparisons between the variables used for the morphological and molecularcharacterization of the seven orchid mycorrhizal fungi isolates using the Tocher clustermethod.

markers, which should lead to the standardization of globaltaxonomical studies of Rhizoctonia species.

Acknowledgements

This work was supported by grants from Fundação deAmparo à Pesquisa do Estado de Minas Gerais (project708/02) and Conselho Nacional de DesenvolvimentoCientífico e Tecnológico (CNPq). We kindly thank Dr. Rob-ert W. Barreto and Dr. Marisa V. de Queiroz for helpful andconstructive suggestions; Dr. Acelino C. Alfenas for assis-tance and information; Dr. Júpiter I. Muro-Abad andMr. Rodrigo B. Rocha for important technical and statisticaladvice; and Dr. Tânia M.F. Salomão for valuable assistancewith the PCR–RFLP analysis. We gratefully acknowledgeDr. R. Larry Peterson, University of Guelph, for his com-ments, helpful suggestions, and advice.

References

Andersen, T.F. 1990. A study of hyphal morphology in the formgenus Rhizoctonia. Mycotaxon, 37: 25–46.

Andersen, T.F. 1996. A comparative taxonomic study of Rhizoc-tonia sensu lato employing morphological, ultrastructural andmolecular methods. Mycol. Res. 100: 1117–1128.

Andersen, T.F., and Rasmussen, H.N. 1996. The mycorrhizal spe-cies of Rhizoctonia. In Rhizoctonia species: taxonomy, molecu-lar biology, ecology, pathology, and disease control. Edited byB. Sneh, S. Jabaji-Hare, S. Neate, and G. Dijst. Kluwer Aca-demic Publishers, Dordrecht, The Netherlands. pp. 379–390.

Arditti, J., Ernst, R., Yan, T.W., and Glabe, C. 1990. The contribu-tions of orchid mycorrhizal fungi to seed germination: a specu-lative review. Lindleyana, 5: 249–255.

Batty, A.L., Dixon, K.W., Brundrett, M., and Sivasithamparam, K.2001a. Contraints to symbiotic germination of terrestrial orchidseed in a Mediterranean bushland. New Phytol. 152: 511–520.

Batty, A.L., Dixon, K.W., Brundrett, M., and Sivasithamparam, K.2001b. Long-term storage of mycorrhizal fungi and seed as atool for the conservation of endangered Western Australian ter-restrial orchids. Aust. J. Bot. 49: 619–628.

Bayman, P., Lebrón, L., Tremblay, R.L., and Lodge, D.J. 1997.Variation in endophytic fungi from roots and leaves ofLepanthes (Orchidaceae). New Phytol. 135: 143–149.

Carling, D.E., Pope, E.J., Brainard, K.A., and Carter, D.A. 1999.Characterization of mycorrhizal isolates of Rhizoctonia solanifrom an orchid, including AG-12, a new anastomosis group.Phytopathology, 89: 942–946.

Cruz, C.D. 1997. Programa GENES: aplicativo computacional emgenética e estatística. UFV, Viçosa, Brazil.

Cruz, C.D., and Regazzi, A.J. 1994. Modelos biométricos aplica-dos ao melhoramento genético. Imprensa Universitária UFV, Vi-çosa, Brazil.

Cubeta, M.A., and Vilgalys, R. 1997. Population biology of theRhizoctonia solani complex. Phytopathology, 87: 480–484.

Cubeta, M.A., Echandi, E., Abernethy, T., and Vilgalys, R. 1991.Characterization of anastomosis groups of binucleate Rhizoc-tonia spp. using restriction analysis of an amplified ribosomalRNA gene. Phytopathology, 81: 1395–1400.

Cubeta, M.A., Vilgalys, R., and Gonzalez, D. 1996. Molecularanalysis of ribosomal RNA genes in Rhizoctonia fungi. InRhizoctonia species: taxonomy, molecular biology, ecology, pa-thology, and disease control. Edited by B. Sneh, S. Jabaji-Hare,

S. Neate, and G. Dijst. Kluwer Academic Publishers, Dordrecht,The Netherlands. pp. 81–86.

Currah, R.S. 1986. Thanatephorus pennatus sp. nov. isolated frommycorrizal roots of Calypso bulbosa (Orchidaceae) from Al-berta. Can. J. Bot. 65: 1957–1960.

Currah, R.S., and Sherburne, R. 1992. Septal ultrastructure of somefungal endophytes from boreal orchid mycorrhizas. Mycol. Res.96: 583–587.

Currah, R.S., and Zelmer, C.D. 1992. A key and notes for the gen-era of fungi with orchids and a new species in the genusEpulorhiza. Rep. Tottori Mycol. Inst. 30: 43–59.

Currah, R.S., Smreciu, E.A., and Hambleton, S. 1990. Mycorrhizaeand mycorrhizal fungi of boreal species of Plantanthera andCoeloglossum (Orchidaceae). Can. J. Bot. 68: 1171–1181.

Currah, R.S., Zettler, L.W., and McInnis, T.M. 1997. Epulorhizainquilina sp. nov. from Platanthera (Orchidaceae) and a key toEpulorhiza species. Mycotaxon, 61: 338–342.

Davidson, R.W., Campbell, W.A., and Blaisedell, D.J. 1938. Dif-ferenciation of wood-decay fungi by their reactions on gallic ortannic acid medium. J. Agric. Res. 57: 683–695.

Dhingra, O.D., and Sinclair, J.B. 1995. Basic plant pathologymethods. CRC Press, Boca Raton, Fla.

Duncan, S., Barton, J.E., and O’Brien, P.A. 1993. Analysis of vari-ation in isolates of Rhizoctonia solani by random amplifiedpolymorphic DNA assay. Mycol. Res. 97: 1075–1082.

Flentje, N.T. 1956. Studies on Pellicularia filamentosa (Pat.) D. P.Rogers – I. Formation of the perfect stage. Trans. Br. Mycol.Soc. 39: 343–356.

Harvais, G., and Hadley, G. 1967. The relation between host andendophyte in orchid mycorrhiza. New Phytol. 66: 205–215.

Jumpponen, A., and Trappe, J.M. 1998. Dark septate endophytes:a review of facultative biotrophic root-colonizing fungi. NewPhytol. 140: 295–310.

Kristiansen, K.A., KjØller, R., Rasmussen, H.N., and Rosendahl,S. 2001. Identification of orchid mycorrhizal fungi fromDactylorhiza majalis (Orchidaceae) based on PCR, SSCP andsequencing of mitochondrial ribosomal LsDNA from singlepelotons. Mol. Ecol. 10: 2089–2093.

Leach, C.M. 1962. Sporulation of diverse species of fungi undernear-ultraviolet radiation. Can. J. Bot. 40: 151–161.

Leake, J. 1994. The biology of myco-heterotrophic (‘saprophytic’)plants. New Phytol. 127: 171–216.

Ma, M., Tan, T.K., and Wong, S.M. 2003. Identification and mo-lecular phylogeny of Epulorhiza isolates from tropical orchids.Mycol. Res. 107: 1041–1049.

Marx, D.H. 1969. Antagonism of mycorrhizal fungi to root patho-genic fungi and soil bacteria. Phytopathology, 59: 153–163.

Mendonça, M.P.,and Lins, L.V. 2000. Lista vermelha das espéciesameaçadas de extinção da flora de Minas Gerais. FundaçãoBiodiversitas & Fundação Zôo-Botânica de Belo Horizonte,Belo Horizonte, Brazil.

Miller, D., and Warren, R. 1995. Some observations on the reintro-duction of orchid species to the high mountain Atlantic rain for-est in the State of Rio de Janeiro. Orchid Rev. 103: 270–273.

Miller, D., Warren, R., and Miller, I. 1996. Orquídeas do Alto daSerra – da mata atlântica pluvial do sudeste do Brasil. EditoraSalamandra Rio de Janeiro, Brazil.

Moore, R.T. 1987. The genera of Rhizoctonia-like fungi:Ascorhizoctonia, Ceratorhiza gen. nov., Epulorhiza gen. nov.,Moniliopsis, and Rhizoctonia. Mycotaxon, 29: 91–99.

Moore, R.T. 1996. The dolipore/parenthesome septum in moderntaxonomy. In Rhizoctonia species: taxonomy, molecular biology,ecology, pathology, and disease control. Edited by B. Sneh, S.

© 2005 NRC Canada

64 Can. J. Bot. Vol. 83, 2005

Jabaji-Hare, S. Neate, and G. Dijst. Kluwer Academic Pub-lishers, Dordrecht, The Netherlands. pp. 13–34.

Mordue, J.E., Banniza, S., Bridge, P.D., Rutherford, M.A., andHolderness, M. 1996. Integrated biochemical, cultural and nu-merical methods. In Rhizoctonia species: taxonomy, molecularbiology, ecology, pathology, and disease control. Edited by B.Sneh, S. Jabaji-Hare, S. Neate, and G. Dijst. Kluwer AcademicPublishers, Dordrecht, The Netherlands. pp. 87–98.

Mori, S.A. 1989. Eastern, extra Amazonian Brazil. In Floristic in-ventory of tropical countries: The status of plant systematic, col-lections, and vegetation, plus recommendations for the future.Edited by D.G. Campbell and H.D. Hammond. The New YorkBotanical Garden, New York. pp. 427–454.

Nei, M., and Li, W.H. 1979. Mathematical model for studying ge-netic variation in terms of restriction endonucleases. Proc. Natl.Acad. Sci. U.S.A. 76: 5269–5273.

Otero, J.T., Ackerman, J.D., and Bayman, P. 2002. Diversity andhost specificity of endophytic Rhizoctonia-like fungi from tropi-cal orchids. Am. J. Bot. 89: 1852–1858.

Pascual, C.B., Toda, T., Raymondo, A.D., and Hyakumachi, M.2000. Characterization by conventional techniques and PCR ofRhizoctonia solani isolates causing banded leaf sheath blight inmaize. Plant Pathol. 49: 108–118.

Pereira, J.M., Barreto, R.W., Ellison, A.C., and Maffia, L.A.2003a. Corynespora casiicola f. sp. lantanae: a potential bio-control agent from Brazil for Lantana camara. Biol. Control,26: 21–31.

Pereira, O.L., Rollemberg, C.L., Borges, A.C., Matsuoka, K., andKasuya, M.C.M. 2003b. Epulorhiza epiphytica sp. nov. isolatedfrom mycorrhizal roots of epiphytic orchids in Brazil. Myco-science, 44: 153–155.

Peterson, R.L., Uetake, Y, and Zelmer, C. 1998. Fungal symbiosiswith orchid protocorns. Symbiosis, 25: 29–55.

Rasmussen, H.N. 2002. Recent developments in the study of or-chid mycorrhiza. Plant Soil, 244: 149–163.

Richardson, K.A., and Currah, R.S. 1995. The fungal communityassociated with the roots of some rainforest epiphytes of CostaRica. Selbyana, 16: 49–73.

Richardson, K.A., Currah, R.S., and Hambleton, S. 1993. Basidio-mycetous endophytes from the roots of neotropical epiphyticorchidaceae. Lindleyana, 8: 127–137.

Roberts, P. 1999. Rhizoctonia-forming fungi: a taxonomic guide.Royal Botanic Gardens, Kew, UK.

Ruschi, A. 1986. Orquídeas do estado do Espírito Santo. EditoraExpressão e Cultura, Rio de Janeiro, Brazil.

Saksena, H.K. 1961. Nuclear phenomena in the basidium ofCeratobasidium praticolum (Kotila) Olive. Can. J. Bot. 39: 717–725.

Sen, R., Hietala, A.M., and Zelmer, C.D. 1999. Common anasto-mosis and internal transcribed spacer RFLP groupings inbinucleate Rhizoctonia isolates representing root endophytes ofPinus sylvestris, Ceratorhiza spp. from orchid mycorrhizas andphytopathogenic anastomosis group. New Phytol. 144: 331–341.

Shan, X.C., Liew, E.C.Y., Weatherhead, M.A., and Hodgkiss, I.J.2002. Characterization and taxonomic placement of Rhizoctonia-like endophytes from orchid roots. Mycologia, 94: 230–239.

Smith, S.E., and Read, D.J. 1997. Mycorrhizal Symbiosis. Aca-demic Press, London.

Sneh, B., and Adams, G.C. 1996. Culture preservation methods formaintaining the genetic integrity of Rhizoctonia spp. InRhizoctonia species: taxonomy, molecular biology, ecology, pa-

thology, and disease control. Edited by B. Sneh, S. Jabaji-Hare,S. Neate, and G. Dijst. Kluwer Academic Publishers, Dordrecht,The Netherlands. pp. 13–35.

Speacht, C.A., Dirusso, C.C., Novotny, C.P., and Ulrich, R.C.1982. A method for extracting high molecular weight deoxyri-bonucleic acid from fungi. Anal. Biochem. 119: 158–163.

Statsoft, Inc. 1995. STATISTICA for Windows. Statsoft, Inc.,Tulsa, Oklahoma.

Stewart, S.L., and Zettler, L.W. 2002. Symbiotic germination ofthree semi-aquatic rein orchids (Habenaria repens, H. quin-quiseta, H. macroceratilis) from Florida. Aquat. Bot. 72: 25–35.

Takahashi, K., Ogiwara, I., and Hakoda, N. 2001. Inoculation ofHabenaria (Pecteilis) radiata with two isolates of orchid mycor-rhizal fungi: germination and growth response to separate, simul-taneous, and consecutive inoculation. Lindleyana, 16: 102–108.

Teather, R.M., and Wood, P.J. 1982. Use of Congo Red-Polysaccharide Interactions in enumeration and characterizationof cellulolytic bacteria from the bovine rumen. Appl. Environ.Microbiol. 43: 777–780.

Tu, C.C., Roberts, D.A., and Kimbrough, J.W. 1969. Hyphal fu-sion, nuclear condition, and perfect stages of three species ofRhizoctonia. Mycologia, 61: 775–783.

White, T.J. Bruns, T.D., Lee, S., and Taylor, J.W. 1990. Amplifica-tion and direct sequencing of fungal ribosomal RNA genes forphylogenetics. In PCR protocols: A guide to methods and appli-cations. Edited by M. Gelfand Maidh, J.J. Sninsky, and T.White. Academic Press, New York. pp. 315–322.

Yang, J., Kharbanda, P.D., Wang, H., and McAndrew, D.W. 1996.Characterization, virulence, and genetic variation of Rhizoctoniasolani AG-9 in Alberta. Plant Dis. 80: 513–518.

Zettler, L.W. 1997a. Terrestrial orchid conservation by symbioticseed germination: techniques and perspectives. Selbyana, 18:188–194.

Zettler, L.W. 1997b. Orchid-fungal symbiosis and its value in con-servation. McIvainea, 13: 40–45.

Zelmer, C.D., and Currah, R.S. 1995. Ceratorhiza pernacatena andEpulorhiza calendulina spp. nov.: mycorrhizal fungi of terres-trial orchids. Can. J. Bot. 73: 1981–1985.

Zettler, L.W., and Hofer, C.J. 1998. Propagation of the little club-spur orchid (Platanthera clavellata) by symbiotic seedgermination and its ecological implications. Environ. Exp. Bot.39: 189–195.

Zettler, L.W., and McInnis, T.M. 1993. Symbiotic germination anddevelopment of Spiranthes cernua and Goodyera pubescens(Orchidaceae: Spiranthoideae). Lindleyana, 8: 155–162.

Zelmer, C.D., Cuthbertson, L., and Currah, R.S. 1996. Fungi asso-ciated with terrestrial orchid mycorrhizas, seeds and protocorms.Mycoscience, 37: 439–448.

Zettler, L.W., Delaney, T.W., and Sunley, J. 1998. Seed propagationof the epiphytic Green-fly orchid, Epidendrum conopseum R.Brown, using its endophytic fungus. Selbayana. 19: 249–253.

Zettler, L.W., Burkhead, J.C., and Marshall, J.A. 1999. Use of amycorrhizal fungus from Epidendrum conopseum to germinateseed of Encyclia tampensis in vitro. Lindleyana, 14: 102–105.

Zettler, L.W., Sunley, J.A., and Delaney, T.W. 2000. Symbioticseed germination of an orchid in decline (Platanthera integra)from the Green Swamp, North Carolina. Castanea, 65: 207–212.

Zettler, L.W., Stewart, S.L., Bowles, M.L., and Jacobs, K.A. 2001.Mycorrhizal fungi and cold-assisted symbiotic germinationof the federally threatened eastern praire fringed orchid,Platanthera leucophaea (Nuttall) Lindley. Am. Midl. Nat. 145:168–175.

© 2005 NRC Canada

Pereira et al. 65